Jökull - 01.12.1968, Qupperneq 34

tion between the size of the open waters and
meteorological and hydrological factors. The
most important parameters seem to be the heat
loss and the river discharge.
Direct observations of the rate of heat loss
from rivers are very limited as yet. We have
considered three methods for such observations.
With the first method the mass of ice in an
ice jam is compared to calculated ice produc-
tion.
In 1965 the volume of an ice jam below
Thjofafoss Falls (Burfell ice jam) in Thjorsa
River was determined with photogrammetric
methods on Febr. 23 and March 27. The in-
crease in ice volume during this period was
19 ± 1 million cubic metres. The mean water
equivalent of the ice jam is not known but on
the surface it was 0.5—0.6. The measured in-
crease in volume occurred in the period March
18—27 and there was no discharge of ice from
the downstream end of the jam in this period.
The total discharge of the river was 124 million
cubic metres March 18—27. The calculated
total ice production in the open water area
upstream from the jam during this period was
15 million metric tons and it is estimated that
about 13 million tons reached the jam. To
bring this into agreement with the measured
increase in volume the mean water equivalent
of the jam must be 0.65—0.72 which is rather
high but not quite unreasonable as the major
part of the jam is dry and the thickness up
to 15 metres.
With the second method the river tempera-
ture is measured in successive sections when the
river is in a state of cooling and the heat loss
is calculated from equation (18).
This method has been tried in Thjorsa River
but has not been sucessful as yet. The diffi-
culties arise from the fact that there are com-
paratively great temperature differences in the
same cross section and there are some practical
obstacles to obtain cross-sectional means. The
Thjorsa River is wide and shallow where the
temperature measurements were made and the
temperature changes in these reaches can be
very rapid as is seen from the following
example from Oct. 13—14 1967. The river width
is about 400 m and the depth 0.5 m, almost
even in a reach of several kilometres. The
discharge was 177 m3 sec-1.
368 JÖKULL 18. ÁR
Time ‘w N vo ew ea
hours °C 0-8 °C m sec-1 mb mb
18 1.09 1.0 1.5 11.5 6.7 5.3
19 0.98 1.2 1.1 10.9 6.5 5.2
20 0.81 2.0 0.9 12.9 6.4 5.0
21 0.60 2.7 0.8 11.9 6.3 5.0
22 0.38 2.3 0.6 12.8 6.2 5.0
23 0.17 1.3 0.2 12.6 6.1 4.8
24 0.0 1.0 -0.1 11.7 6.1 4.5
01 0.0
Here N, ta, V6, ew and ea are hourly means.
The water temperature was zero at 2350 hours.
It is interesting that in this example the air
temperature is a little higher than the water
temperature and accordingly there is a slight
heat gain by convection but the heat loss by
radiation and evaporation is many times higher.
The water was slightly supercooled (tw rz -^0.01
°C) at about the same time as the air tempera-
ture was down to 0 °C and frazil ice produc-
tion had started.
The third method consists in measuring the
frazil ice discharge in successive sections of the
river and obtain the heat loss from the open
water area from the increase in frazil ice dis-
charge. This would be the most rational ap-
proach where the heat loss formulas are used
for calculations of ice production as it is likely
that the heat loss is affected by clusters of
frazil ice floating on the surface. An ice dis-
charge gauge based on the fact that the elec-
trical conductivity of the ice-water mixture is
a function of the ratio between ice volume
and total volume of the mixture has been
constructed by Mr. Björn Kristinsson, Elec-
tronics, Reykjavik, but it has been on the ex-
perimental stage up to now.
CONCLUDING REMARKS
Calculations of heat loss from a river sur-
face and frazil ice production on the basis of
meteorological observations are possible, pro-